A History Of Violence: The Impact Of Early Violence Exposure On Financial Risk Preferences

James Byder

Universidad EAFIT – School of Economics and Finance – Center for Research in Economic & Finance (CIEF)

Diego Agudelo

Universidad EAFIT; Universidad EAFIT – School of Economics and Finance – Center for Research in Economic & Finance (CIEF)

Mateo Uribe Castro

Universidad EAFIT

September 2, 2015

Center for Research in Economics and Finance (CIEF), Working Papers, No. 15-24


This paper examines whether growing up in areas with high homicide rates affects financial risk preferences. Our key conjecture is that individuals who have grown up in violent areas possess more risk averse financial preferences. We find support for this hypothesis using a unique dataset of mutual fund investors from one of Colombia’s largest stock brokers alongside Colombian official data on homicide rates.

A History Of Violence: The Impact Of Early Violence Exposure On Financial Risk Preferences – Introduction

Does being raised in a violent environment affect individual’s attitudes towards risk? We seek answers to this question by studying the investment decisions of a large number of Colombian mutual fund investors. Our paper is one of the first to examine the behavior of mutual fund investors in Latin America with Colombia providing an ideal setting in which to study the impact of violence.

Being a witness or victim of violence is regarded by mental health organizations as one of the most traumatic experiences an individual can undergo (World Health Organization 2013). It is well documented in the clinical psychology literature that traumatic experiences lead to physical and mental health problems (see e.g. McNally 2003, Boscarino 2006). Given their powerful effect on our emotions, there is reason to believe traumatic experiences may also affect our preferences towards risk. This potentially important determinant of risk preferences combined with the large number of people worldwide affected by traumatic events has fuelled a growing interest from both economists and policy makers. The resulting literature has sought to connect traumatic experiences with individual’s risk preferences. Such studies have looked at the impact of civil war (see e.g. Kim and Lee 2014; Voors, Nillesen, Verwimp, Bulte, Lensink, and Van Soest 2012), natural disasters (see e.g. Bucciol and Zarri 2015; Cassar, Healy, and Von Kessler 2011; Cameron and Shah forthcoming), and terrorist attacks (see e.g. Sacco, Galletto, and Blanzieri 2003). With the exception of Voors et al. (2012), most studies find that risk aversion increases following traumatic events.

Financial Risk Preferences

The traumatic events literature builds on previous work relating individual risk preferences with their life experiences. A seminal paper of this kind by Malmendier and Nagel (2011) identified that individuals growing up during periods of financial crisis are more risk averse in their investment decisions. Similarly, Knupfer, Rantapuska, and Sarvimaki (2013) looking at evidence from the Finnish Great Depression find that negative labor market experiences increase aversion towards risky assets. More broadly, this strand of literature complements other papers that have highlighted the importance of both demographic and socio-economic factors (see e.g. Booth and Nolen 2012; Hryshko, Luengo-Prado, and Sorensen 2011) and genetics (see e.g. Barnea, Cronqvist, and Siegel 2010; Calvet and Sodini 2013).

In our paper we look again at the importance of experience and measure the impact growing up in violent areas has on an individual’s financial risk preferences. Violence in this context has not been studied before. A recent paper by Callen, Isaqzadeh, Long, and Sprenger (2014) using data from Afghanistan finds that individuals exposed to violence exhibit an increased preference for certainty. However, the violence measured in Callen et al. (2014) derives exclusively from acts of war. We build on Callen et al. (2014) by observing violence in a more natural setting. We take Colombia a country with a troubled and diverse history of violence and measure the impact of early childhood exposure to violence on individuals financial risk preferences. Our approach contributes to the literature for two main reasons. First, the violence we are able to measure offers a more complete account of violence than has been previously studied. This is because war, rather than violence, has been the context of previous papers ( Callen et al. (2014), Voors et al. (2012) and Kim and Lee 2014). War is a transitory phenomenon that contains negative aspects, independent of violence, that may be associated with risk preferences. Such elements include invasion, occupation, destruction, decline in social spending, deterioration in infrastructure, famine, wide spread emigration, genocide and ethnic cleansing. War alters the normal context in which people live and in such an environment it may be difficult to unpick the marginal effect of violence from the for-mentioned elements. Our purpose is this paper is not to study war or the violence associated with it.

Financial Risk Preferences

We want to understand a different type of violence, one that occurs in a natural, every day setting. We believe Colombia is one of the best countries to study for this purpose because its history of violence is unique, stems from a variety of causes and has formed part Colombia’s identity and culture. Our second contribution is that we are able to observe individual’s actual investment decisions, whereas previous studies in the traumatic events literature have had to rely on experiments or survey data to ascertain risk preferences. This additional evidence is of particular value useful because while the literature on traumatic events broadly supports the argument that risk aversion increases following such experiences, the evidence with respect to war or violence associated with it is more mixed. Two previous papers find risk aversion increases (Callen et al. (2014), Kim and Lee 2014) while a third finds that it falls(Voors et al. (2012)). Although we are measuring everyday violence and not the effect of war our paper fits most closely with this part of the traumatic events literature. Our results come down firmly on the side of increased risk aversion bringing the balance of evidence with respect to violence related events in line with the more general traumatic experiences.

Financial Risk Preferences

See full PDF below.